U.S. patent application number 11/938035 was filed with the patent office on 2009-05-14 for control channel detection scheme.
Invention is credited to Jens Berkmann, Mauro Bottero.
Application Number | 20090122891 11/938035 |
Document ID | / |
Family ID | 40561010 |
Filed Date | 2009-05-14 |
United States Patent
Application |
20090122891 |
Kind Code |
A1 |
Bottero; Mauro ; et
al. |
May 14, 2009 |
CONTROL CHANNEL DETECTION SCHEME
Abstract
A method for detection of a control channel includes receiving
data transmitted via the control channel. A control channel receive
quality is estimated based on a metric difference between a metric
of a known final trellis state and a minimum metric amongst the
metrics of the trellis states based on the received data. It is
decided whether or not to detect the control channel depending on
the estimated control channel receive quality.
Inventors: |
Bottero; Mauro; (Mougins Le
Haut, FR) ; Berkmann; Jens; (Muenchen, DE) |
Correspondence
Address: |
Thomas G. Eschweiler, Esq.;Eschweiler & Associates, LLC
Suite 1000, 629 Euclid Avenue
Cleveland
OH
44114
US
|
Family ID: |
40561010 |
Appl. No.: |
11/938035 |
Filed: |
November 9, 2007 |
Current U.S.
Class: |
375/265 |
Current CPC
Class: |
H03M 13/658 20130101;
H03M 13/4107 20130101; H03M 13/41 20130101; H04L 1/0072 20130101;
H04L 1/0054 20130101 |
Class at
Publication: |
375/265 |
International
Class: |
H04L 5/12 20060101
H04L005/12 |
Claims
1. A method for detection of a control channel, comprising:
receiving data transmitted via a control channel; estimating a
control channel receive quality based on a metric difference
between a metric of a known final trellis state and a minimum
metric amongst the metrics of the trellis states, the metrics
associated with the received control channel data; and deciding
whether or not to detect the control channel based on the estimated
control channel receive quality.
2. The method of claim 1, wherein the control channel receive
quality comprises a ratio of the metric difference and another
metric difference, the another metric difference being associated
with the minimum metric and a maximum metric amongst the metrics of
the trellis states.
3. The method of claim 2, wherein the control channel receive
quality is related to the expression
R=(M.sub.0-M.sub.min)/(M.sub.max-M.sub.min), wherein M.sub.0 is the
metric of the known final trellis state, M.sub.min is the minimum
metric amongst the metrics of the trellis states and M.sub.max is
the maximum metric amongst the metrics of the trellis states.
4. The method of claim 1, wherein deciding whether or not to detect
the control channel comprises: comparing the control channel
receive quality with a threshold value; and deciding to detect the
control channel if the control channel receive quality is higher
than the threshold value.
5. The method of claim 4, further comprising: deciding to not
detect the control channel if the control channel receive quality
is below the threshold value.
6. The method of claim 1, wherein data transmitted via a plurality
of control channels is received, further comprising: estimating the
control channel receive qualities of the plurality of control
channels; comparing the computed control channel receive qualities
with each other to determine the control channel having the maximum
control channel receive quality; and deciding whether or not to
detect the control channel having the maximum control channel
receive quality.
7. The method of claim 6, wherein the control channels of the
plurality of control channels are high speed shared control
channels of a wireless communications system.
8. A receive unit for detecting a control channel, comprising: a
receiver configured to receive data transmitted via a control
channel; a channel quality estimator configured to compute a
control channel receive quality based on a metric difference
between a metric of a known final trellis state and a minimum
metric amongst the metrics of the trellis states, wherein the
metrics are associated with the received control channel data; and
a deciding unit configured to decide whether or not to detect the
control channel based on the estimated control channel receive
quality.
9. The receive unit of claim 8, wherein the channel quality
estimator is configured to compute the control channel receive
quality as a ratio of the metric difference and another metric
difference, wherein the another metric difference is associated
with the minimum metric and a maximum metric among the metrics of
the trellis states.
10. The receive unit of claim 9, wherein the channel quality
estimator is configured to compute
R=(M.sub.0-M.sub.min)/(M.sub.max-M.sub.min), wherein R is related
to the control channel receive quality, and wherein M.sub.0 is the
metric of the known final trellis state, M.sub.min is the minimum
metric amongst the metrics of the trellis states, and M.sub.max is
the maximum metric among the metrics of the trellis states.
11. The receive unit of claim 8, wherein the deciding unit
comprises: a threshold comparator configured to compare the
computed control channel receive quality with a threshold value,
wherein the deciding unit is further configured to decide to detect
the control channel if the control channel receive quality is
higher than the threshold value.
12. The receive unit of claim 11, wherein the deciding unit is
further configured to decide not to detect the control channel if
the control channel receive quality is below the threshold
value.
13. The receive unit of claim 8, wherein the receiver is configured
to receive data transmitted via a plurality of control channels and
the channel quality estimator is configured to compute control
channel receive qualities of the plurality of control channels,
further comprising: a quality comparator configured to compare the
computed control channel receive qualities with each other to
determine the control channel having the maximum control channel
receive quality, and wherein the deciding unit is configured to
decide whether or not to detect this control channel having the
maximum control channel receive quality.
14. The receive unit of claim 8, wherein a first section of the
channel quality estimator which is configured to compute the metric
difference is implemented in hardware.
15. The receive unit of claim 8, wherein a second section of the
channel quality estimator which is configured to compute the
control channel receive quality based on the metric difference is
implemented in software.
16. The receive unit of claim 8, wherein the deciding unit is
implemented in software.
17. A method for selection of a control channel from a plurality of
control channels, comprising: receiving data transmitted via the
plurality of control channels; estimating a control channel receive
quality based on a metric difference between a metric of a known
final trellis state and a minimum metric amongst the metrics of the
trellis states for each of the plurality of control channels,
wherein the metrics are associated with the received data of each
respective control channel; and selecting one of the plurality of
control channels based on an evaluation of the control channel
receive qualities.
18. The method of claim 17, wherein the control channel receive
quality comprises a ratio of the metric difference and another
metric difference, the another metric difference being associated
with the minimum metric and a maximum metric amongst the metrics of
the trellis states.
19. The method of claim 18, wherein the control channel receive
quality is related to the expression
R=(M.sub.0-M.sub.min)/(M.sub.max-M.sub.min), wherein M.sub.0 is the
metric of the known final trellis state, M.sub.min is the minimum
metric amongst the metrics of the trellis states, and M.sub.max is
the maximum metric amongst the metrics of the trellis states.
20. The method of claim 18, wherein selecting one of the plurality
of control channels depends on an evaluation of the control channel
receive qualities further comprises: comparing the computed control
channel receive qualities with each other to determine the control
channel having the maximum control channel receive quality; and
selecting that control channel having the maximum control channel
receive quality.
21. A receive unit for selecting a control channel from a plurality
of control channels, comprising: a receiver configured to receive
data transmitted via the plurality of control channels; a channel
quality estimator configured to compute a control channel receive
quality based on a metric difference between a metric of a known
final trellis state and a minimum metric amongst the metrics of the
trellis states for each control channel, wherein the metrics are
associated with the received data for each respective control
channel; and a selector configured to select a control channel
depending on an evaluation of the computed control channel receive
qualities.
22. The receive unit of claim 21, wherein the channel quality
estimator is configured to compute the control channel receive
quality as a ratio of the metric difference and another metric
difference, the another metric difference being associated with the
minimum metric and a maximum metric amongst the metrics of the
trellis states.
23. The receive unit of claim 22, wherein the channel quality
estimator is configured to compute
R=(M.sub.0-M.sub.min)/(M.sub.max-M.sub.min) for each of the
plurality of control channels, wherein R is related to the control
channel receive quality, wherein M.sub.0 is the metric of the known
final trellis state, M.sub.min is the minimum metric amongst the
metrics of the trellis states, and M.sub.max is the maximum metric
amongst the metrics of the trellis states.
24. The receive unit of claim 22, wherein the selector comprises: a
metrics comparator configured to compare the computed control
channel receive qualities with each other to determine the control
channel having the maximum control channel receive quality.
Description
FIELD OF THE INVENTION
[0001] The invention relates generally to the technique of control
channel detection in communications systems.
BACKGROUND OF THE INVENTION
[0002] In many communications systems, in particular wireless
mobile communications systems, one or more control channels are
transmitted in addition to data channels. Such a control channel
may contain information which must be known at the receiver before
starting the detection of the data channel. Therefore, a fast
detection of the control channel at a receiver is important for
obtaining a high overall system performance.
BRIEF DESCRIPTION OF THE DRAWINGS
[0003] Aspects of the invention are made more evident by way of
example in the following detailed description of embodiments when
read in conjunction with the attached drawing figures, wherein
[0004] FIG. 1 is a schematic block diagram of a decoding and
detection unit of a receiver according to a first embodiment;
[0005] FIG. 2 is a schematic timing diagram of shared control
channels and data channels;
[0006] FIG. 3 is a schematic illustration of a shared control
channel encoding process;
[0007] FIG. 4 is a schematic block diagram of a decoding and
detection unit of a receiver according to a second embodiment;
[0008] FIG. 5 is a graph illustrating an average bit error rate
after re-encoding and an average control channel quality for a
control channel which is not intended for a receiver; and
[0009] FIG. 6 is a graph illustrating an average bit error rate
after re-encoding and an average control channel quality for a
control channel which is intended for a receiver.
DETAILED DESCRIPTION OF THE INVENTION
[0010] In the following, embodiments of the invention are described
with reference to the drawings, wherein like reference numerals are
generally utilized to refer to like elements throughout the
description. In the following description, for purposes of
explanation, numerous specific details are set forth in order to
provide a thorough understanding of one or more aspects of
embodiments of the invention. It may be evident, however, to one
skilled in the art that one or more aspects of the embodiments of
the invention may be practiced with a lesser degree of these
specific details. In other instances, known structures and devices
are shown in a simplified representation in order to facilitate
describing one or more aspects of the embodiments of the invention.
The following description is therefore not to be taken in a
limiting sense, and the scope of the invention is defined by the
appended claims.
[0011] In mobile communications systems, a transmitter transmits
user data via user data channels and control data via one or more
control channels to a receiver. The transmitter may be a base
station of the wireless communications system and the receiver may
be a mobile station of the wireless communications system. Such
channels from a base station to a mobile station are termed
downlink channels. However, the following description may also
relate to data channels and control channels for transmission of
user data and control information, respectively, transmitted from a
mobile station to a base station. Such channels are referred to as
uplink channels. Herein and in the related art, a mobile station
will be referred to as a User Equipment (UE).
[0012] Wireless communications systems according to the following
description may be CDMA (Code Division Multiple Access) systems in
one embodiment. However, also other types of multiple access
techniques could be used in wireless communications systems
considered herein, and all such variations and alternatives are
contemplated as falling within the scope of the invention.
[0013] A control channel intended for a receiver may contain an
Identity (ID) bit sequence of the scheduled receiver. Usually, the
ID bit sequence is contained in a leading part of the control
channel transmitted by the transmitter. For instance, in this
leading part of the control channel, the ID bit sequence may be
used to mask additional control information such as e.g.
information relating to the channelization code and/or the
modulation scheme used at the transmitter to generate the user data
signal to be transmitted. Therefore, the detection of the control
channel may be based on an estimation of the decoding quality of
the control channel at the receiver. If the control channel can be
decoded with high (or better to say sufficient) quality, it may be
assumed that the leading part of the control channel had been
masked by the ID bit sequence of the scheduled receiver.
[0014] The detection performance of a receiver may be defined in
terms of the "probability of missed detection" (P.sub.md), i.e. the
probability that the receiver misses the detection of the part of
the control channel carrying its ID bit sequence, and the
"probability of false alarm" (P.sub.fa), i.e. the probability that
the receiver falsely detects a part of the control channel carrying
a different receiver's ID bit sequence or any other bit,
sequence.
[0015] Typically, the detection of a control channel is a crucial
algorithm for communications system performance. Missed detections
reduce the communications system throughput and false detections
trigger unnecessary receptions which will be aborted at a later
stage of the reception process. However, this causes waste of
terminal resources and power consumption.
[0016] FIG. 1 illustrates a first embodiment of a decoding and
detection unit 100 of a receiver in a wireless communications
system, which may be a UE or a base station. In the following
embodiment, without limitation of the generality, the receiver is
assumed to be part of a UE (i.e. a mobile station).
[0017] The decoding and detection unit 100 of a receiver contains a
first section 110 and a second section 120. The first section 110
comprises a de-masking unit 111, a decoder 112, a minimum/maximum
unit 113, and a difference unit 114. Control data contained in a
downlink control channel and user data contained in a downlink data
channel are input to the de-masking unit 111 via input 115. The
de-masking unit 111 further receives the ID of the UE under
consideration. The part of the control sequence which is masked by
the ID of the scheduled UE is de-masked by the ID of the UE under
consideration. De-masking may be performed by the inverse masking
process, e.g. by a sign inversion process based on the UE's ID.
Control data and user data are supplied by a demodulator (not
shown) of the receiver to the input 115 of the first section
110.
[0018] The de-masked part of the control sequence is fed via
connection 116 to the decoder 112. In one embodiment the decoder
112 may be a Viterbi decoder performing channel decoding. Further,
the decoder 112 may perform a de-puncturing operation.
[0019] The Viterbi decoder 112 decodes the incoming data, here at
first the de-masked leading part of the control channel. As known
in the art, Viterbi decoders 112 are used to decode bit sequences
which were subjected to channel encoding at the transmitter.
[0020] Viterbi decoding is based on finding the shortest path
through a state diagram of an encoder register which is used for
channel encoding at the transmitter. This diagram is known as a
trellis diagram. In the trellis diagram, the states of the encoder
register are plotted versus the discrete time k. According to the
Viterbi algorithm, a branch metric which represents a measure of
the probability of the branch is calculated for each possible
branch between two states (previous state relating to the time
stamp k.fwdarw.destination state relating to the time stamp k+1).
The branch metrics are then added to the respective state metrics
(which are frequently also referred to as path metrics in the
literature) of the previous states (ADD). For branches leading to
the same destination state, the sums which are obtained in this way
are compared (COMPARE). That branch to the destination state under
consideration whose sum of the branch metric and state metric of
the previous state is a minimum is selected (SELECT) and forms the
extension of the path leading to this previous state to the
destination state. These three basic operations of the Viterbi
algorithm are known as ACS-(ADD-COMPARE-SELECT) operations.
[0021] While from a combinational point of view, the number of
paths through the trellis diagram increases exponentially as k
rises (that is to say as time progresses), it remains constant for
the Viterbi algorithm. This is because of the selection step
(SELECT). Only one selected path ("survivor") per destination state
is retained and can be continued. The other possible paths are
rejected. Recursive path rejection is the concept used in the
Viterbi algorithm to limit the number of paths while progressing
through the trellis diagram.
[0022] In the following, without limitation of generality, it may
be assumed that the length of the channel encoder register at the
transmitter is n. This means that the encoder register has n
register cells. In this case, the encoder register may be described
by a trellis diagram having 2.sup.n states. Thus, the decoder 112
has 2.sup.n outputs 117, wherein at each output one state metric
(of the 2.sup.n state metrics) is output and updated every time
stamp k.
[0023] It is now considered a time stamp k.sub.dec at which time
the part of the control channel containing the de-masked ID of the
UE is completely decoded. This is achieved when the bits of the
de-masked ID are all trellis processed and when the encoder's state
is known. The latter condition (known encoder's state) may be
guaranteed by forcing the encoder to a zero state by appending an n
bit sequence of zeros to the masked part of the control sequence
containing the ID of the scheduled UE. By way of example, if n=8, a
tail bit sequence (0,0,0,0,0,0,0,0) forces the channel encoder to
its zero state. As a consequence, the state metric associated with
the zero state in the trellis diagram at the Viterbi decoder 112
should be the minimum state metric amongst all state metrics at
time stamp k.sub.dec, which is n time stamps later than the time
stamp corresponding to the last bit of the ID.
[0024] In the following, M.sub.0 refers to the state metric at the
known final trellis state (e.g. at the zero trellis state if a zero
tail sequence is used), M.sub.min refers to the minimum state
metric amongst the state metrics of the trellis states and
M.sub.max refers to the maximum state metric amongst the state
metrics of the trellis states at the 2.sup.n state metric outputs
of the decoder 112 at time stamp k.sub.dec. M.sub.min and M.sub.max
are determined by the minimum/maximum unit 113 in one
embodiment.
[0025] The difference unit 114 calculates the metric differences
M.sub.0-M.sub.min and, optionally, M.sub.max-M.sub.min. These
differences are passed via connections 118, 119 to a computing unit
121 located in the second section 120 of the receiver 100. In one
embodiment the computing unit 121 may compute the following ratio
of metric differences
R = M 0 - M min M max - M min . ##EQU00001##
R provides a fair indication of the detection quality of the
control channel at the UE (or, more generally, at the receiver). R
may assume values in the range between 0 and 1, and the decoding
quality improves as R gets close to 0. More specifically, as the
minimum metric state should be the forced zero state, an error-free
Viterbi decoding should result in R=0. Due to noise, it may be the
case that the forced zero state is not the state having minimum
state metric, i.e. M.sub.min<M.sub.0. However, also in this
case, the zero state metric MO should at least be close to the
minimum one M.sub.min. Therefore, R stays small. On the other hand,
when the Viterbi decoding has a substantial number of errors (for
instance for a control channel which is not masked with the ID of
the UE under consideration), the zero state metric M.sub.0 can even
be closer to the maximum state metric M.sub.max. In this case, the
ratio R increases indicating that the control channel is detected
with low quality. Thus, the control channel receive quality is
related to R and may thus be evaluated on the basis of R.
[0026] It is to be noted that the metric difference
M.sub.max-M.sub.min is used for scaling the metric difference
M.sub.0-M.sub.min. The scaling is beneficial because Viterbi state
metrics increase with the amplitude of the received signal, i.e.
the SNR (Signal-to-Noise Ratio). Scaling by M.sub.max-M.sub.min
eliminates this influence of signal amplitude to the effect that R
is virtually independent of the amplitude of the received signal.
Further, it is known that Viterbi decoders such as e.g. real
fixed-point Viterbi decoders scale the accumulated state metrics
during the forward recursion, for instance by subtracting a
constant from all the state metrics almost every time stamp in
order to avoid overflow. Also such internal re-scaling operation of
the Viterbi decoder 112 is canceled out by considering a ratio of
metric differences.
[0027] It is to be noted that in alternative embodiments the
denominator of the ratio R may be of a different type, i.e. need
not necessarily be a difference M.sub.max-M.sub.min. For instance,
also an average metric could be used. In this case, however, R
would not be limited by 1.
[0028] R is passed via connection 122 to a quality comparator 123.
The quality comparator 123 compares R with a threshold value T. T
may e.g. be a fixed threshold value in one embodiment. If R is
below or equal to the threshold value T, the control channel is
decided to be a control channel of sufficient detection quality.
Then, the detected control information (e.g. the modulation scheme
and/or the channelization code) is output from the second section
120 of the receiver at control information output 124 and the
reception of the user data channel is triggered at a first control
output 125 enabling user data detection. Otherwise, if R is above
the threshold value T, the detected control information is
discarded (in FIG. 1, this is illustrated by an open switch 126
between the Viterbi decoder 112 and the control information output
124) and the detection of the user data channel is inhibited by
activating a control output 127 which disables the detection of the
user data. Thus, it is deciding to detect the control channel (i.e.
to use its control information) whenever the detection quality
(which may be expressed e.g. by R.sup.-1) is higher than a
threshold (in this case T.sup.-1).
[0029] According to one embodiment the first section 110 of the
receiver may be implemented in dedicated hardware whereas the
second section 120 of the receiver may be implemented in software,
e.g. as a General Purpose Processor (GPP). This solution provides
for competitional efficiency because in such alternative
embodiments all subtractions are performed in hardware, and for
high flexibility because the threshold T and probably further
algorithms for deciding about the detection of a control channel
may be programmable by software.
[0030] In FIGS. 2 to 6 a specific embodiment relating to High Speed
Downlink Packet Access (HSDPA) is shown. HSDPA has been introduced
in the third Generation Partnership Project (3GPP) Release 5 to
provide enhanced support for packet data services with improved
system throughput and reduced system latency for peak data rates up
to 14.4 Mb/s in the downlink direction from the base station to the
UE. Most of the description of the embodiment illustrated in FIG. 1
may equally be applied to the second embodiment and is therefore
partly omitted in order to avoid reiteration. On the other hand,
details of the more specific second embodiment may equally apply to
related aspects of the first embodiment.
[0031] HSDPA supports enhanced features such as shared channel
transmission, adaptive modulation and coding (AMC), fast Hybride
Automatic Repeat reQuest (HARQ), fair and fast scheduling at the
NodeB (i.e. base station) rather than at the radio network
controller (RNC) and Fast Cell Side Selection (FCSS). Further,
HSDPA introduces a shorter Transmission Time Interval (TTI) of 2 ms
(corresponding to 3 time slots) than in the previous releases in
order to decrease Round Trip Time (RTT) delay. Amongst others,
HSDPA introduces new shared and fast-scheduled physical channels:
[0032] HS-PDSCH (High Speed Physical Downlink Shared Channel)
carries user data in the downlink direction. It is time shared
between the UEs (i.e. mobile stations). To achieve a peak transfer
rate of 14.4 Mb/s the NodeB can allocate up to 15 HS-PDSCHs to the
same UE. [0033] HS-SCCH (High Speed Shared Control Channel) is used
by the NodeB to signal to the scheduled UE to receive HS-PDSCH(s)
in the next TTI. In the meantime, it carries the HS-PDSCH(s)
control information such as channelization codes, modulation
schemes (e.g. QPSK (Quadrature Phase Shift Keying) or 16 QAM
(Quadrature Amplitude Modulation)), transport block size, HARQ
process number, redundancy and constellation version and a new data
indicator. Further, the HS-PDSCH(s) control information contains an
identity (ID) of the UE to which the message is addressed. In order
to schedule several UEs, the NodeB can transmit up to 4 HS-SCCHs
simultaneously.
[0034] FIG. 2 illustrates the timing of control channels HS-SCCHs
and user data channels HS-PDSCHs. HS-SCCH(s) are sent two time
slots in advance the corresponding HS-PDSCH(s) to allow enough time
to the UE to configure itself for receiving the data channels. A
first HS-SCCH 201 is addressed to a first UE denoted as UE1 and a
second HS-SCCH 202 is addressed to a second UE denoted as UE2. Each
HS-SCCH 201, 202 is divided into two functional parts: in Part 1,
which spans over one time slot, the NodeB transmits urgent
information such as the channelization codes and the modulation
schemes while in Part 2, which spans over two time slots, the
remaining (less time critical) information is transmitted. In the
example illustrated in FIG. 2, two HS-SCCHs 201, 202 are
transmitted simultaneously to UE1 and UE2. Two time slots later,
i.e. at a time instance in the center of Part 2, UE1 and UE2
simultaneously start to detect user data channels HS-PDSCHs. More
specifically, UE1 detects three HS-PDSCHs and UE2 detects four
HS-PDSCHs during a first TTI of 2 ms (3 slots) referred to as TTI1.
For the next transmission time interval TTI2, only control
information in form of a HS-SCCH 203 for UE2 is provided. Thus,
during TTI2, only UE2 is active and detects 7 HS-PDSCHs. For TTI3,
the NodeB only supplies control information in form of a HS-SCCH
204 for UE1. Therefore, during TTI3, only UE1 is active and uses 5
HS-PDSCHs. This process of providing demodulation information by
HS-SCCHs dedicated to a specific UE and demodulating the
corresponding HS-PDSCHs in UE1 and/or UE2 continues during
subsequent transmission time intervals TTI4, TTI5.
[0035] FIG. 3 illustrates one embodiment of HS-SCCH encoding. The
abbreviations used in FIG. 3 are: [0036] CCS: Channelization Code
Set (7 bits) [0037] MS: Modulation Scheme (1 bit) [0038] TBS:
Transport Block Size (6 bits) [0039] HAP: Hybrid-ARQ Process (3
bits) [0040] RV: Redundancy and constellation version (3 bits)
[0041] ND: New date indicator (1 bit) [0042] CRC: Cyclic Redundancy
Check (16 bits) [0043] UE ID: User Equipment Identity (16 bits) A
HS-SCCH comprising Part 1 having 40 bits and Part 2 having 80 bits
is denoted by reference number 301. As already mentioned HS-SCCH
301 spans over three times slots. In Part 1, CCS and MS information
302 are convolutionally encoded and masked (by an XOR operation)
with the encoded 16-bits ID 303 of the scheduled UE (referred to as
UE ID). In Part 2, TBS, HAP, RV and ND information 304 are
convolutionally encoded together with a Cyclic Redundancy Check
(CRC) masked with the UE ID. The same 1/3-rate convolutional code
with 256 states is used in Part 1 and Part 2 encoding. Puncturing
is used to reduce the number of bits of the convolutionally encoded
UE ID information 303, convolutionally encoded CSS and MS
information 302 and convolutionally encoded TBS, HAP, RV, ND, CRC
and UE ID information 304, 305 to reduce the number of bits. Up to
4 HS-SCCHs addressing different UEs can be simultaneously
transmitted from the serving HS-DSCH NodeB. Thus, each UE has to
monitor all 4 HS-SCCHs in order to allow for a fast detection of
the use of its UE ID in Part 1 in each HS-SCCH 301. Further, in
connection with the embodiments described herein, the content of
3GPP TS 34.121 Release 5 is incorporated by reference into this
document.
[0044] Thus, referring also to FIG. 2, after receiving Part 1 of
all the HS-SSCH(s), the UE has only one time slot to make the
decision: if one of the HS-SSCH(s) employing its UE ID is detected,
the UE should configure the reception of the HS-PDSCH(s) with the
decoded CCS and MS of the detected HS-SCCH.
[0045] FIG. 4 illustrates a decoding and detection unit 400 of a
receiver according to a second embodiment. This receiver is adapted
to provide for a fast detection of HS-SCCH(s) in HSDPA or, more
generally, in a mobile communications system providing a plurality
of shared control channels. The decoding and detection unit 400
comprises first sections 410.1, 410.2, . . . , 410.4 which may be
implemented in dedicated hardware and a second section 420 which
may be implemented in software e.g. as a GPP. The first sections
410.1-410.4 are configured to detect the four shared control
channels HS-SCCH#1-HS-SCCH#4. The sections 410.1-410.4 may be
implemented in one hardware instance which can be multiplexed in
time according to the number of monitored control channels
according to one embodiment. Instead, the plurality of first
sections 410.1-410.4 may be also implemented in parallel hardware.
The second section 420 is typically implemented in one GPP which is
fed by the e.g. time multiplexed outputs of the four first sections
410.1-410.4.
[0046] The implementation and operation of each first section
410.1-410.4 are similar to the implementation and operation of the
first section 110 illustrated in FIG. 1, and the description
thereof is analogously applicable to first sections 410.1-410.4.
Each first section 410.1-410.4 comprises a de-masking unit 411, a
decoding unit 412, a minimum/maximum unit 413 and a difference unit
414 which correspond to units 111, 112, 113 and 114 of FIG. 1,
respectively. Here the 16-bit UE ID is convolutionally encoded and
punctured in UE processing unit 431 according to the Part 1
generation scheme of a HS-SCCH shown in FIG. 3. A bit sequence
which is coming from a demodulator arranged in the signal path
upstream of the first sections 410.1-410.4 is input via input
connection 415.1 to the de-masking unit 411. There, it is de-masked
by sign inversion dependent from the output of the UE processing
unit 431.
[0047] Similar to the first embodiment, the example viterbi decoder
412 provides a final accumulated metric for each trellis state at
the end of the forward recursion performed for decoding Part 1 of
the HS-SCCH 301. Referring to FIG. 3, both the UE ID information
303 and the CCS and MS information 302 are terminated by a 8-bit
tail sequence 303t. This 8-bit tail sequence may be
(0,0,0,0,0,0,0,0), thus forcing the 8-bit register of the channel
encoder to the zero state at the end of encoding the Part 1 of
HS-SCCH 301. Of course, any other state known to the receiver could
be chosen, and such alternatives are contemplated by the invention.
As state metrics computed by a Viterbi decoder 112, 412 may reflect
the Euclidian distance between the received bit sequence and the
bit sequence corresponding to the most likely path through the
trellis diagram, the metric of the zero state should be the minimum
metric amongst all metrics computed by the Viterbi decoder 412 at
the end of the Part 1 decoding forward recursion. Therefore, as
already explained further above, the metric difference
M.sub.0-M.sub.min is a measure of the quality of detection of the
HS-SCCH under consideration. In order to avoid fixed-point problems
(e.g. an overflow) during Viterbi decoding, the state metrics are
typically scaled during the forward recursion by subtracting a
constant to all state metrics every or almost every time stamp. For
example, the minimum branch metric computed at each time stamp may
be subtracted from all state metrics. It is to be noted that such
scaling is typically different for decoding different HS-SCCHs,
i.e. among decoder units 412 which decode different control
channels.
[0048] The difference units 414 output metric differences
M.sub.0-M.sub.min and M.sub.max-M.sub.min for each HS-SCCHs. These
metric differences are fed into computing units 421.1-421.4, which
are implemented in software in one embodiment. In other words, the
second section 420 (e.g. a GPP) computes in a time multiplexed
fashion ratios R for all HS-SCCHs which are presently detected.
[0049] The ratios R associated with the qualities of the detected
HS-SCCHs are input to a control channel selection unit 428. The
control channel selection unit 428 may be implemented by a minimum
algorithm selecting the ratio R which is smaller than all other
ratios R delivered to the control channel selection unit 428. This
ratio R.sub.min is passed to a quality comparator 423 having the
same functionality as quality comparator 123 of the first
embodiment. Further, a HS-SCCH index denoting the HS-SSCH having
the minimum R (i.e. R.sub.min) is output from the control channel
selection unit 428 and passed to a selector 426.
[0050] The operation of the second section 420 is similar to the
operation of the second section 120 illustrated in FIG. 1, and the
description thereof is analogously applicable to second section
420. Briefly, if R.sub.min is greater than the (e.g. fixed)
threshold T, the detection quality of all HS-SCCHs is too poor and
no detection of a HS-PDSCH is initiated. In this case, a control
output 427 is activated. Otherwise, if R.sub.min is below or equal
to the threshold T, a control output 425 (corresponding to control
output 125 of the first embodiment) is enabled and reception of the
corresponding HS-PDSCH is initiated. In this case, the selector 426
is activated by the quality comparator 423 and the CCS and MS
information contained in Part 1 of the selected HS-SCCH indexed by
the control channel selection unit 428 is provided at output 424 of
the second section 420. On the basis of this information, the
receiver configures itself to start user data reception from the
HS-PDSCH associated with the selected HS-SCCH one time slot
later.
[0051] It is to be noted that the scaling of the metric difference
M.sub.0-M.sub.min by the metric difference M.sub.max-M.sub.min in
the denominator of ratio R in one embodiment allows to compare the
ratios R obtained from different HS-SCCHs in the control channel
selection unit 428, because this decoder-individual scaling further
eliminates the use of different scaling factors in Viterbi decoding
of different HS-SCCHs.
[0052] The software-hardware split illustrated in FIGS. 1 and 4
should be understood to be exemplary. For instance, another
embodiment possibility is to implement the minimum-maximum
operation performed by the minimum/maximum units. 113, 413 and the
subtraction operation performed by the difference units 114, 414 in
software. In this case, the first section 110, 410.1-410.4 may be
implemented by standard hardware and substantially all computation
for control channel quality estimation is done in software. In
another embodiment, the computing units 121, 421.1-421.4 and/or
others of the units in the second sections 120, 420 may be
implemented in hardware.
[0053] The second embodiment provides for a reliable and fast
HS-SCCH detection without adding more complexity, i.e. hardware
modules, to the receiver. The proposed HS-SCCH detection algorithm
is based on the Viterbi decoders final state metrics which are
exploited at the end of the Part 1 decoding. This approach provides
for a finer tuning of the threshold and therefore for a better
P.sub.md and P.sub.fa joint optimization compared to a conventional
HS-SCCH detection and selection approach which is based on a BER
(Bit Error Rate) channel quality estimation of HS-SCCHs. Such BER
quality estimation involve re-encoding of the decoded data stream
and comparing the re-encoded data with the received data prior to
Viterbi decoding. If re-encoded data is substantially the same as
the data before Viterbi decoding, the corresponding HS-SCCH is
rated to have a high detection quality. Otherwise, the detection
quality of the. HS-SCCH is rated to be low and the data detection
of the corresponding user data channel HS-PDSCH is inhibited. FIGS.
5 and 6 illustrate graphs which compare this conventional approach
("average BER after re-encoding") with the approach described
herein ("average R"). Average BER after re-encoding and average R
are plotted versus Eb/NO (the energy per bit to noise power
spectral density ratio) in units of dB. FIG. 5 shows the results
when a HS-SCCHs not intended for a UE is detected and FIG. 6 shows
the results when a HS-SCCH is detected which is intended for the
UE. The conventional approach suffers from low sensitivity due to
the limited number of bits available before channel decoding. In
contrast thereto, using average R discrimination, a higher
sensitivity and thus a considerably better P.sub.md and P.sub.fa
joint optimization is possible.
[0054] It is to be noted that the detection of one or more control
channels as described above is applicable to a wide range of
receivers, including 3GPP-receivers, HSDPA-receivers, LTE-receivers
(Long Term Evolution) etc. Further, the concepts described above by
way of example in relation to downlink channels are also applicable
to uplink control channels and user data channels, i.e. to a
receiver located in a base station. For instance, HSUPA-receivers
(High Speed Uplink Packet Access) may be implemented that way for
enhanced detection of one or more uplink control channels
transmitted by UEs.
[0055] Although the invention has been illustrated and described
with respect to one or more implementations, alterations and/or
modifications may be made to the illustrated examples without
departing from the spirit and scope of the appended claims. In
particular regard to the various functions performed by the above
described components or structures (assemblies, devices, circuits,
systems, etc.), the terms (including a reference to a "means") used
to describe such components are intended to correspond, unless
otherwise indicated, to any component or structure which performs
the specified function of the described component (e.g., that is
functionally equivalent), even though not structurally equivalent
to the disclosed structure which performs the function in the
herein illustrated exemplary implementations of the invention. In
addition, while a particular feature of the invention may have been
disclosed with respect to only one of several implementations, such
feature may be combined with one or more other features of the
other implementations as may be desired and advantageous for any
given or particular application. Furthermore, to the extent that
the terms "including", "includes", "having", "has", "with", or
variants thereof are used in either the detailed description and
the claims, such terms are intended to be inclusive in a manner
similar to the term "comprising".
* * * * *